opportunities to strengthen the medicare advantage star ratings program€¦ · • meaningful use...

Opportunities to Strengthen the Medicare Advantage Star Ratings Program

December 2017

Public Policy Institute


2Opportunities to Strengthen the Medicare Advantage Star Ratings Program

Contents

Overview 3

Understanding MA Star Ratings 5

Study Findings 10

Recommendations 21

Conclusion 24

Appendix A 25

Endnotes 27

• The Medicare Advantage (MA) Star Ratings system is one of a number of quality incentive programs used to link Medicare program payments to quality performance.

• While the Star Ratings program has been successful in encouraging quality improvement across MA plans, the program can be strength-ened to better incentivize plan improvements.

• More closely aligning the MA Star Ratings with other Medicare quality incentive programs will help improve performance across Medicare.

KEY HIGHLIGHTS



Overview

There has been a growing emphasis among policymakers, government agencies, employers, and health plans on paying for healthcare based on the quality of care delivered rather than only on the quantity of services provided.

There are many types of quality incentive programs, which can include things such as pay-for-reporting or pay-for-performance programs, bonus payments, payment penalties and value-based purchasing (VBP) programs. These programs are intended to drive spending in healthcare toward higher value and more efficient providers and services. In particular, policymakers have focused on expanding the use and reach of pay-for-performance programs in Medicare, especially following enactment of the Affordable Care Act (ACA), which included a number of provisions that introduced new quality incentive programs or strengthened existing ones.1

One of the quality incentive programs enacted under the ACA is the Medicare Advantage (MA) Star Ratings (“MA Star Ratings”) program which links MA plans’ payments to their performance on quality measures. The MA Star Ratings focus on measuring the quality of private plans providing the Medi-care benefit to enrollees under Medicare Part C, and usually Part D as well.

There are also a number of quality incentive programs used in fee-for-service (FFS) Medicare applicable to providers such as hospitals, physicians, home health agencies (HHAs), and dialysis centers, to name a few. The FFS programs range in type from pay-for-reporting of quality measures to more robust VBP programs. We examined the MA Star Ratings program in the context of a number of other Medicare FFS quality incentive programs as shown in Table 1 (see Appendix A for an overview of Quality Incentive Programs discussed in this paper).

Table 1 Medicare Quality Incentive Programs Reviewed in this Study

Provider/Plan Type Quality Incentive Programs Reviewed

Medicare Advantage • MA Star Ratings (including Part D measures)

Hospitals* • Hospital Inpatient Quality Reporting (IQR) and Outpatient Quality Reporting (OQR)

• Hospital Readmissions Reduction Program (HRRP)

• Hospital Acquired Condition (HAC) Reduction Program

• Hospital Valued Based Purchasing (VBP)

Accountable Care Organizations (ACOs) • ACO Quality Measure Program

Physicians • Physician Quality Reporting System (PQRS)

• Meaningful Use

• Physician Value-Based Payment Modifier (VBPM)/ Merit-Based Incentive Payment System (MIPS)

Dialysis Facilities • End Stage Renal Disease (ESRD) Quality Improvement Program (QIP)

Home Health Agencies • Home Health Agency (HHA) Quality Reporting

• HHA Pilot VBP

Skilled Nursing Facilities (SNFs) • Skilled Nursing Facilities (SNF) VBP

* There are additional quality incentive programs which may impact hospital payments, such as for inpatient psychology units, which were not included in this paper.

Policymakers are increasingly focused on expanding the use and reach of quality incentive programs in Medicare.



Quality incentive programs can be structured in a variety of ways. The programs examined for this report fall into one of five types: pay-for- reporting, payment penalties, bonus payments, VBP programs, or shared savings eligibility (see Table 2).

In some instances more than one type of incentive program may apply to a particular provider group. For example, inpatient hospitals are subject to pay-for-reporting requirements (Inpatient Quality Reporting (IQR)), payment penalties (Hospital Readmissions Reduction Program (HRRP) and Hospital Acquired Conditions (HAC) Program), as well as a VBP program (Hospital VBP). A number of the pay-for-performance programs or VBP programs evolved from longer-standing pay-for-reporting requirements. CMS, consistent with the direction in the ACA and the Medicare Access and CHIP Reauthorization Act (MACRA), is moving away from simply incentivizing reporting to using quality information to tie payment more closely to higher quality performance.

MA plans receive a bonus to their base payment levels for higher quality performance levels. Despite a mostly upside system for MA plans in terms of quality performance, it should be noted that there is also an inherent downside from a competitive perspective (such as less money for extra benefits to attract beneficiaries). MA plans that consistently perform poorly on the MA Star Ratings can be removed from program participation, so there is a very real negative consequence for poor quality performance by plans.

Table 2 Types of Quality Incentive Programs in Medicare

Type of Program Description Examples

Pay-for-Reporting A portion of payments are withheld unless providers report quality measures or fulfill other programmatic requirements

• PQRS

• IQR and OQR

• HHA Quality Reporting

Payment Penalties Payments are reduced for covered entities whose quality performance on certain measures falls below specified thresholds or is low relative to the group

• HRRP

• HAC Reduction Program

• ESRD QIP

• Meaningful Use

Bonus Payments Covered entities are rewarded through greater payments for achieving higher quality performance

• MA Star Ratings

VBP Programs Payments may be increased, decreased, or unchanged based on quality performance. VBP programs withhold a portion of payments from all entities and then redistribute the withheld payments based on performance

• Hospital VBP

• Physician VBPM/MIPS

• SNF VBP

• HHA Pilot VBP

Shared Savings Requirement Organizations are not able to share in savings achieved through better care management if they do not meet quality thresholds

• End Stage Renal

* There are additional quality incentive programs which may impact hospital payments, such as for inpatient psychology units, which were not included in this paper.

CMS is moving away from simply incentivizing reporting to using quality information to tie payment more closely to higher quality performance.



Payment Implications

In general, the quality incentive programs reviewed affect 1 percent to 5 percent of payments for any given provider (see Appendix A for program specific details). Some programs like the Hospital VBP and the Physician VBPM had payment implications phased in to their eventual levels, which are 2 percent for the Hospital VBP in 2017 and thereafter2 and 2 percent to 4 percent for the Physician VBPM in 2017, depending on the physician practice group size3 (beginning in 2019 physicians will be paid under the Merit-Based Incentive Payment System (MIPS)).4

As discussed above, it is important to note that some provider groups are affected by multiple programs and therefore have a greater total share of their payments impacted by quality programs. For example, 6 percent of inpatient hospital payments are tied to performance, which does not include the penalties associated with not reporting on required quality measures.5 Quality scores also impact ACO payments, preventing shared savings payments if the ACO does not report the measures or perform at sufficient quality levels. Finally, as described below in more detail, MA plans that achieve 4-5 Stars see significant revenue boost—their payment benchmarks are increased by 5 percent to 10 percent of county level FFS costs.

Understanding MA Star Ratings

In the MA program, private health plans provide the Medicare benefit to enrollees as an alternative to traditional Medicare, or Medicare FFS.

Although plans are required to cover all Medicare services, they have some flexibility in how they design their benefits, they may limit choice of providers to those in a plan’s network, and they can offer additional benefits beyond FFS. The ACA required significant changes to MA plan payments, including linking MA plan payments to the plan’s quality rating, or MA Star Rating.

MA Plan Payments Are Tied to MA Star Ratings

MA plans are paid based on administratively set county-specific payment rates (or “benchmarks”). The ACA aligned the MA county level benchmarks more closely with FFS costs in each county. Plans submit bids representing their estimated costs for providing the Medicare Parts A and B benefits and the MA plan bids are compared to the benchmarks. If a plan bids below the benchmark, it retains a portion of the difference between its bid and the benchmark, which is known as a “rebate.” Rebates can be used to provide extra benefits for enrollees such as reduced cost sharing or enhancing the



Part D benefit if the plan offers prescription drug coverage. If a plan bids above the benchmark, it must charge a premium to the enrollee for the amount above the benchmark. Under the ACA, higher performing plans—those with 4 or more Stars—are eligible for higher county benchmarks, and plans with 3.5 Stars or more are eligible for larger rebates.

The MA Star Ratings use a five-star quality rating system to measure MA plan performance. As mentioned above, there are two ways plans are rewarded based on their Star Ratings. The first is through increases to the benchmarks. In most counties, plans with 4-5 Stars receive a bonus in the form of an increase to their county benchmark equal to 5 percent of the county’s underlying FFS costs (meaning plans bid against a higher amount and are therefore eligible for larger rebates if they bid below the benchmark or do not have to charge as large a premium if they bid above the benchmark). In some counties, known as double bonus counties, the benchmark is increased by 10 percent of underlying FFS costs. However, because the ACA limited benchmarks to no more than they would have been prior to the ACA, in some counties, this “benchmark cap” prevents plans from receiving the full amount of the quality bonus they earned.

Second, MA Star Ratings determine the rebates that plans receive when they bid below the benchmarks. Prior to the ACA, MA plans received a rebate equal to 75 percent of the difference between the plan’s bid and the bench-mark. The ACA created a tiered system for rebate amounts based on the Star Ratings as shown below, with the highest performing plans retaining the largest portion of the difference.

Table 3 MA Performance Based Payment Changes

Plan Star Rating* Benchmark Increase Rebate Percentage

4.5 to 5 Stars 5% of county FFS (10% in double bonus counties) 70%

4 Stars 5% of county FFS (10% in double bonus counties) 65%

3.5 Stars N/A 65%

3 or Less Stars N/A 50%

* New plans from new parent organizations receive a benchmark increase equal to 3.5% of county FFS costs and 65% rebate amount.



Determination of the MA Star Ratings

MA plans are evaluated on a variety of measures, each of which is also rated on a five-star scale. Plans then receive an overall score based on weighted performance across the measures. For MA plans that provide Part D coverage (also known as MA-PDs), the overall score includes performance on both Part C and Part D measures. For 2017, there are 32 Part C measures and 15 Part D measures. MA-PD plans’ overall scores reflect performance on 44 measures; three measures overlap and are included in both the MA and Part D measure sets but are included only once in the overall MA-PD score.

The MA Star Ratings include different categories of measures. These categories are process measures, beneficiary experience measures, outcome and intermediate outcome measures, access measures and improvement measures. The measures are differentially weighted in the score calculation to emphasize the importance of some measures (see Table 4). For example, outcome measures have three times the contribution to the overall score as process measures. The improvement measures—used to evaluate year- to-year changes in performance—have the greatest weight (5 times the process measures).

In order to determine the star rating for each of the individual measures, CMS first compares the “raw performance” across all plans on each measure for a year and looks for statistically significant “cutpoints.” 6 These cutpoints determine the scores needed to achieve each level of stars (e.g., 3 Stars, 4 Stars). In this way, for most measures, MA plans are rated on a performance curve and the actual scores needed to achieve a particular Star Rating can change annually.

Once a plan’s measure-specific Star Ratings are determined, CMS calculates the improvement measure scores. Inclusion of improvement measures depend on the plan’s scores without the improvement measures in part to hold higher performing plans harmless for minor changes in performance.7 Finally, CMS calculates the overall score, using the measure weighting approach described above. Finally, CMS calculates the overall score, using the measure weighting approach described above.

Table 4 MA Star Ratings Measure Categories

Measure Category Weight in Summary and Overall Scores

Process 1

Beneficiary Access or Beneficiary Experience and Complaints 1.5

Intermediate Outcome to Outcome 3

Improvement 5



Following the overall score calculation, CMS makes two additional adjust-ments to plans’ scores. First, CMS will apply a reward factor for plans that demonstrate consistently high performance across a number of measures. The range of factors applied for high performance is 0.1 to 0.4 stars. This adjustment can have a meaningful impact on the final MA Star Rating.

Finally, beginning in 2016, CMS started to employ a categorical adjustment index (CAI) to adjust plans’ MA Star Ratings to account for the effects of serving higher proportions of low income or disabled beneficiaries. However, the actual rating adjustment amounts for the contracts with higher propor-tions of beneficiaries with low incomes or disabilities has been very small and for many contracts is not enough to move their overall scores. In 2017, only 15 plan contracts (or 4 percent of plan contracts) saw a change to their star ratings because of the CAI and the remaining 349 plan contracts (96 percent) saw no change. Therefore, despite the significant complexity and resource intensity of this approach, it is expected to have relatively little impact—changing the scores of only a small proportion of plans.8

The MA Star Ratings affect payments nearly a year and a half after they are posted to CMS’ Plan Finder website as a tool for beneficiaries shopping for coverage. So, for example, the 2017 Star Ratings, which are published on Plan Finder in October 2016 for the 2017 annual enrollment period, are based on performance data collected in 2015 and 2016 and are then used to adjust 2018 payments, as Figure 1 shows. CMS proposes and finalizes the Star Ratings metrics and methodology in 2015-2016, while data are being collected.

Figure 1 2017 Star Ratings: From Measure Announcement to Plan Payment

The MA Star Ratings affect payments nearly a year and a half after they are posted to CMS’ Plan Finder website as a tool for beneficiaries shopping for coverage.

CMS notes forecasted changes for the 2017 MA Star ratings.November 2014

CMS previews 2017 MA Star Ratings measuresand methodology for plan comments. November 2015

CMS proposes final 2017 MA Star Ratings measures and methodology. Stakeholdershave 30 day comment opportunity.

February 2016

CMS finalizes 2017 MA Star Ratings measuresand methodology.April 2016

CMS announces the 2017 plan ratings and posts them on Medicare Plan Finder ahead of 2016 annual enrollment period.

October 2016

CMS adjusts 2018 plan payments based on 2017 MA Star Ratings.

January – December2018

Measure Announcement

Data Collection for 2017 MA Star Ratings

2017 MA Ratings

2017 MA Ratings Impact 2018 Payment



MA Plan Quality Is Improving

In MA, data from CMS on the distribution of contracts’ Star Ratings from 2013 to 2017 illustrates gains in quality among MA plans. In 2017, nearly 49 percent of MA-PD contracts are 4-5 Star contracts, 1.75 times higher than in 2013 when only 28 percent of contracts had 4-5 Stars. Additionally, there are only 12 contracts that are below 3 Stars in 2017, compared to 62 in 2013. Perhaps more importantly, in 2017, approximately 68 percent of MA-PD enrollment was in contracts with 4-5 Stars compared to 2013, when only 38 percent of enrollees were in higher performing plans (see Figure 2). As plans have improved their performance, it appears that enrollees are also increasingly selecting higher quality MA plans.

Figure 2 Distribution of MA-PD Enrollees by Contract Quality Score, 2013-2017

About 68%of MA enrollees are in 4-5 Star plans in 2017, up from 38% in 2013.

2013 20152014 2016 2017

0

5%

10%

15%

20%

25%

30%

35%

40%

2 Stars 2.5 Stars 3 Stars 3.5 Stars 4 Stars 4.5 Stars 5 Stars

Source: Centers for Medicare & Medicaid Services. 2017 Star Ratings.



Study Findings: Comparing Medicare Quality Incentive Programs

Measure Sets

The number of measures included in the MA Star Ratings could be reduced

The number and types of quality measures used in the different programs reviewed vary widely. Some of the programs used by CMS are more limited in scope and focus on improving quality in a specific area, such as readmissions or hospital-acquired conditions, and therefore have only one or a few measures. Other programs evaluate providers or plans over a wider array of quality metrics. The inclusion of too many measures in a quality incentive program may undermine quality improvement efforts by limiting the ability to target improvements, diluting the importance of individual measures, and makes it more challenging for stakeholders (including consumers) to understand drivers of quality.9

Penalty programs like HRRP, HAC and ESRD QIP tend to have fewer mea-sures with 6, 6 and 11 measures each, respectively, in 2017. At the other end of the spectrum, reporting programs tend to include larger sets of measures, with over 271 measures for MIPS in 2017, and over 50 measures for IQR. However, it is important to differentiate these reporting programs, since any one provider does not need to report on all measures. Rather, providers select and report on a much smaller set of applicable measures. The number of measures included in Medicare VBP programs tends to be more moderate, with 14 measures included in the Hospital VBP in 2017 and for MIPS in 2017, physicians are required to report at least 6 measures, including at least one outcome measure.

ACOs, perhaps the model most similar to MA plans in terms of the overall responsibility for different aspects of patient care, report on 33 total measures. However, for scoring purposes these 33 measures are rolled up to 23 measures (6 of the CAHPS measures are combined into one measure and there are composite measures for diabetes and coronary artery disease).

The MA Star Ratings program includes the greatest number of measures for plans to report and be evaluated on among VBP programs. MA-PD plans are evaluated on 44 measures across Parts C and D. Additionally, as mentioned, CMS also collects and reports on plan performance on measures that are not part of the Star Ratings program, but may either be under consideration for inclusion in the measure set or may be retired from the measure set, on the Display Page. While not directly included in the MA Star Ratings, plans also have to focus attention on the large number of measures that are included on the Display Page.

The MA Star Ratings program includes the greatest number of measures (44) for plans to report and be evaluated on among VBP programs.

44



While the volume of metrics means that no particular measure takes on too much importance and a wider variety of processes and outcomes can be included, having too many measures can dilute plans’ ability to target performance improvement to those areas that are most important and clinically meaningful.

Many quality incentive programs experience significant annual changes in measures and methodology

CMS has introduced numerous annual changes to quality incentive programs, which not only require considerable resources from plans and providers to follow, monitor, and comply with each year, but also have the potential to destabilize year-to-year performance. These annual changes include things such as:

• Adding or retiring multiple measures in a given year;

• Adding or changing the weights of measures or measure categories;

• Removing or changing performance thresholds; and

• Changing the way improvement is measured and/or incorporated into the rating methodology.

As a result, these changes in methodology and measure sets can have meaningful implications for payments even absent no real changes in performance.

As an example, the Hospital VBP program has had relatively large changes to the number of measures included, the categories of measures included, and the weighting of categories of measures since it was first implemented (FY 2013).10 A recent Government Accountability Office (GAO) report examined trends in payment adjustments and quality performance under the Hospital VBP program.11 GAO found that while most hospitals had payment adjustments (both positive and negative) of less than 0.5 percent of applicable Medicare payments each year, the share of hospitals receiving adjustments of greater than 0.5 percent increased each year, growing from 7 percent in 2013 to 26 percent in 2015. Importantly, the GAO did not find real changes in the trends of quality performance due to the program. However, changes in the score calculations and weighting of measures may have changed the performance scores even when there was little change in actual quality.

MA plans have also experienced notable annual changes in the MA Star Ratings measures and methodology. For example, between 2011 and 2012, CMS retired 10 measures (nearly 20 percent of the total measure set in 2011) and added 10 new measures for 2012. Additionally, CMS introduced the measure weighting system that same year—prior to that the measures had not been weighted.

Changes in methodology and measure sets can have meaningful implications for payments even absent no real changes in performance.



It is clear that while CMS continues to focus on evolving the quality incentive programs used in Medicare for the future, it will also continue to make annual changes to these programs. It is important that the MA Star Ratings program continue to provide a sufficient level of transparency and input (including formal comment opportunities), and that all changes be implemented on a prospective, rather than retrospective, basis.

Outcome measures remain underrepresented in many measure sets including the MA Star Ratings

Across Medicare quality programs, CMS has consistently expressed an interest in focusing on outcome measures. Outcome measures are preferred by CMS because they measure changes in a beneficiary’s health as a result of the care that is provided to them, as compared to process measures which only evaluate whether certain activities take place.12 Outcomes measures are more likely to assess quality performance on measures of clinical relevance to Medicare beneficiaries.

CMS emphasizes outcome measures in quality incentive programs in one of two ways. The first is by increasing the number of outcome or intermediate outcome measures included in the measure sets. The second way, and perhaps the more relied upon method today, is through use of a weighting methodol-ogy that values outcome measures more than process or other types of measures. Part of the need to use weighting to increase the value of outcome measures, rather than simply adding more outcome measures, is that there is currently a limited number of outcome measures available for inclusion.

Despite an increasing focus on outcome measures or a move to make the measures more clinically relevant, process measures still tend to significantly outnumber the other types of measures in Medicare’s quality incentive programs. For example, the 2017 MIPS Quality measure set includes 271 measures total, of which 182 (67 percent) were process measures and only 73 (27 percent) were outcome or intermediate outcome measures (the remaining measures were efficiency measures, a patient engagement and experience measure, and structure measures). For the 2017 MA Star Ratings for MA-PDs, the number of process measures (16) was nearly double the number of intermediate outcome (6) and outcome (3) measures combined. However, because of the weighting methodology CMS employs, the outcome and intermediate outcome measures account for approximately 34 percent of the Star Ratings, even though the number of outcome measures account for just 21 percent of the total measures.

Process measures still tend to significantly outnumber the other types of measures in Medicare’s quality incentive programs.



The table below shows the number of measures by type in the MA Star Ratings as well as the weighted contribution to the overall score.

Outcomes measures are especially important in moving quality measurement toward clinically meaningful measures. Despite the efforts to include more outcome measures or to value them more heavily in score calculations, these measures–which assess actual improvements in beneficiary health status—are still underrepresented.

Quality incentive programs do not adequately account for differences due to socioeconomic status

A key issue for a number of quality incentive programs is whether the quality measurement system sufficiently addresses differences in performance due to plans or providers serving higher proportions or disproportionate numbers of beneficiaries who are lower income or have lower socioeconomic status (SES). Plans or providers serving higher proportions of beneficiaries with lower socioeconomic status may have distinct challenges in quality performance that may lower their performance comparatively. Under quality incentive programs, the potential to remove resources from plans and providers that disproportionately serve these populations and may already receive lower reimbursements may mean they have even less resources to reinvest in quality improvement.

Ideally, measures—or the ultimate ratings—impacted by socioeconomic factors would be adjusted to better account for the influence of SES on performance. However, many measures and rating systems do not yet account for SES and, as a result, likely disadvantage plans or providers that serve these populations. While the National Quality Forum (NQF) has recently revised their position on adjusting measures for SES, stating that some measures may need adjustments, this is a relatively new position and has not been widely adopted by CMS or measure developers yet.13

Table 5 MA Measure Type as Percent of Total Measures, Unweighted and Weighted

MA Star Rating Measure Type 2017 Measure Count

Percent of Total Measures

Weighted Measure Value*

Weighted Measure Percent of Total Weight

Process 16 36% 16 20%

Access 7 16% 10.5 13%

Experience 10 23% 15 19%

Intermediate Outcome 6 14% 18 23%

Outcome 3 7% 9 11%

Improvement 2 5% 10 13%

Source: Centers for Medicare & Medicaid Services. Medicare 2017 Part C & D Star Rating Technical Notes.

* Note this does not reflect that new measures all receive a weight of 1 their first year no matter what type.



Further, the Department of Health and Human Services (HHS) Assistant Secretary for Planning and Evaluation (ASPE) recently published a report on the impact of low SES on Medicare’s VBP programs, including MA.14

This report clearly describes the significant impact of social risk factors on quality ratings and VBP across the Medicare program. Two key findings in this report were:

1. Beneficiaries with social risk factors had worse outcomes on many quality measures, regardless of the providers they saw, and dual enrollment status was the most powerful predictor of poor outcomes; and

2. Providers that disproportionately served beneficiaries with social risk factors tended to have worse performance on quality measures, even after account-ing for their beneficiary mix.

Specific to MA, the ASPE report found that MA contracts with a high proportion of members with social risk factors generally did worse on overall quality scores, and were much less likely to receive quality bonus payments. The findings also suggested that the CAI did not sufficiently address the issue for plans and there needed to be a more significant and longer term adjustment.15

In the meantime, one possible approach CMS could take is to make the adjustment amounts under the CAI more meaningful to impacting overall Star Rating scores.

The lack of an SES adjustment for most measures included in the hospital quality incentive programs, particularly under the HRRP program, has drawn continued criticism and concern.16 Numerous studies have demonstrated that hospital readmission rates are influenced by the SES characteristics of the population that the hospital serves.17 Studies have also demonstrated that safety net hospitals tend to have higher readmission rates.18 Additionally, there have been similar concerns raised about the Hospital VBP program.19 One recent review of the hospital pay-for- performance programs found that hospitals most likely to be safety net hospitals (those receiving the highest amount of disproportionate share payments) were two times more likely to receive a payment penalty under the VBP compared to hospitals less likely to be safety net hospitals.20

Similarly, the need for adjustments to physician quality measures has also been raised. In recent NQF guidance on the measures under consideration for the MIPS program, one of the overarching themes was the need to better account for the impact of patients’ SES on quality performance.21 Specifically, they noted that “[t]he impact of patients’ SES and other demographic factors on measure results should continue to be explored, and it is important to take into account whether providers are caring for high-risk populations.”22



The lack of an adjustment for SES for many of the MA and Part D Star Ratings measures has also been a serious concern for many MA plans. While some of the measures, like CAHPS measures, are case mix adjusted, other measures are not adjusted to account for the SES or disabled status of enrollees. There have been a number of studies demonstrating that plans that serve a higher proportion of low SES enrollees tend to have lower overall Star Ratings.23

In response to concerns raised by plans, CMS conducted a study which showed a moderate within contract impact to performance for certain measures resulting from SES and disability status.24 CMS has reached out to measure developers to see if measure adjustment is warranted, but in the meantime, CMS continues to make the interim adjustment known as the CAI to the MA Star Ratings scores to account for differences in performance based on their research on within contract differences for certain measures.25 However, as noted above, the CAI has a very limited effect for most MA plans and most plans are concerned that it does not adequately address the issue.

Scoring Methodology

Quality incentive programs’ methodologies are often overly complicated

In some of the quality incentive programs reviewed, the methodology for calculating the quality scores is quite complicated, raising questions about whether this complexity creates disincentives for participation (where voluntary) or confusion as to how the overall scores are computed. Further, the complexity in assigning scores can attenuate the connection between measure-level improvements and overall quality score. As a result, the burden associated with participation may not be commensurate with the payment incentives.

A number of factors contribute to the complexity of a scoring methodology, including the number of measures included, how performance scores are assigned for the measures, how the overall score is calculated including any weighting methodology, and how adjustments to scores (e.g., to account for low SES or to reward high performing plans or providers) can impact the final performance scores.

As shown in Figure 4, the MA Star Ratings has a particularly complicated methodology for several reasons:

• A large number of measures are included in the ratings, particularly for MA-PD plans.

• The cut points, or thresholds, for assigning Star Ratings to the individual measures are determined annually and after the data are collected, rather than prior to the measurement period.



• Measures receive varying weights in the final scoring, dependent on their type (e.g., process, outcome).

• The final scoring includes a complicated improvement measure, which factors in year-over-year performance on a subset of the measures but holds some plans harmless for consistently high performance.

• Additional adjustments are made before arriving at the final score including the CAI adjustment for differences in low-SES enrollment across contracts and application of a “reward factor” for plans with consistently high scores across multiple measures.

These adjustments add complexity and make it difficult for plans to relate improved performance on any one measure with improvements in their overall Star Ratings scores.

Assign Individual Measure Star Ratings

CMS establishes measure-level Star Ratings for most measures by using either a clustering analysis or relative distribution analysis. This analysis is done annually on all plan scores so that plans are “graded on a curve.”

Calculate Improvement Measures

There are separate Part C and Part D improvement measures for MA-PD contracts. Improvement measures are calculated using only measures that have prior year data.

Weight Individual Measures Based on Measure Weights

CMS then calculates weighted averages of the measure level stars based on whether the measure is a process, outcome or intermediate outcome, patient experience and access, or improvement measure.

Apply Improvement Measures

For most plans, CMS includes the improvement measures in the summary and overall scores. However, if the contract has 2 Stars or fewer, no improvement measure is included and if a contract has more than 4 Stars, CMS takes the higher of the scores with or without the improvement measures.

Apply Rewards Factor

CMS calculates the mean and variance for all measures and categorizes performance to increase scores for consistently high performance.

Apply Adjustment for SES (CAI)

CMS adjusts scores for a subset of measures and categorizes plans by the proportion of low income or disabled enrollees to adjust plans’ scores.

Calculate Final Summary and Overall Ratings Scores

Figure 3 MA-PD Star Ratings Score Calculation Overview



Other programs reviewed had similar issues with complexity. Notably, the physician quality incentive programs (PQRS, VBPM, and Meaningful Use) have been considered extremely complex and onerous and have seen low participation rates by providers. For example, in 2015, 49 percent of physicians still did not participate in reporting on quality through PQRS and 40 percent received negative payment adjustments of 1.5 percent.26 Lack of participation is largely because of the complexity of participation but there are real questions about how much physicians can move quality performance at an individual level, especially under a system they may not fully understand.27 Some of the sweeping changes that were enacted under MACRA were intended, in part, to address the complexity of the current PQRS, meaningful use, and VBPM requirements.

More specifically, MACRA aims to streamline the requirements associated with physician quality incentive programs and to simplify the process to encourage greater participation and engagement. Congress enacted MACRA with two tracks for physician participation—one tied to quality performance under MIPS and the other tied to participation in alternative payment models. Despite the intention of MACRA to simplify current requirements, initial reactions from the provider community suggest that many physicians will find the new requirements as complicated as or more complicated than the current system.28

Another issue with quality measurement complexity is the issue of investing in and developing systems to track performance or assist in identifying areas for performance improvement. While this may be an issue for providers generally, ACOs are also grappling with the complexity of their score devel-opment process and tracking measure performance. For instance, if ACOs do not, or cannot, invest in tracking measures that are claims-based, they may be missing opportunities to improve quality. These complexities may have disadvantaged some ACOs and even led to their disqualification from sharing in savings despite lowering costs.29

Overall, the significant processes involved in assessing quality performance and determining quality scores may make it difficult for plans and providers to effectively monitor and impact performance. In part this is due to the attenuation of measure performance on overall scores, where the more complicated the process the more difficult for individual measure improve-ments to meaningful move a score. Plans and provider must also invest significant time and resources to understanding these quality evaluations.



Unlike other programs, measuring improvement in the MA Star Ratings can penalize overall performance score

The MA Star Ratings include improvement measures that assess plans’ improvements on Part C and Part D measures. However, if plans perform poorly on the improvement measures, they can reduce a plan’s overall Star Rating—this is due in large part to the high weight assigned to the improve-ment measures in the overall Star Ratings.

This situation is unique among the quality incentive programs examined. Other programs reviewed compare measurements of performance achievement to performance improvement and take the higher score. Therefore, improvement is only incorporated into a score if it increases the performance score.

As an example, the Hospital VBP—used as the model for many of the VBP systems in terms of incorporating improvement (SNP VBP, HHA VBP)—compares measure performance to previous year’s performance as well as to other hospitals’ performance. In this way, the Hospital VBP scoring process evaluates both achievement and improvement for each measure and the hospital receives the higher of the two scores—achievement or improve-ment. As a result, the program evaluates improvement on a measure-by- measure basis and rewards providers that have significant improvements, but it does not necessarily penalize providers that did not improve or even declined compared to their baseline performance, if that performance is still favorable relative to other hospitals.

The improvement measures for the MA Star Ratings work differently. The improvement measures for both Part C and Part D compare performance over two years on a number of measures. These comparisons result in the assignment of a Star Rating for each of the Part C and Part D improvement measures. These measures are worth 5 times the process measures and are therefore the most heavily weighted measures in the MA Star Ratings. Therefore, a low rating on the improvement measure(s) could impact up to 13 percent of the plans’ overall Star Rating (see Table 6 above). CMS does apply a “hold harmless” approach for plans that see a reduction in measure scores but still have a 5-Star Rating overall.



Use of pre-set measure thresholds uncommon among quality incentive programs but may have value

While most of the quality incentive programs reviewed do not include pre-set thresholds for performance, many, like the Hospital VBP program, measure improvement by comparing the performance period against a baseline period. In this way, while not necessarily a benchmark of all other facilities’ performance, providers can benchmark against previous performance. And significant improvement from year-to-year can be used as the measure score, if more favorable than performance relative to other facilities’.

Under MACRA, CMS will also provide thresholds for the total composite score. These performance thresholds will be determined annually as the mean or median of the MIPS scores for all eligible providers in a prior period as selected by CMS. Initially, in the first years of the MIPS program, CMS will not have the historical data to develop the performance thresholds but will instead use a combination of performance level on other quality incentive programs including meaningful use.

For MA, CMS used to announce pre-determined four Star performance thresholds for certain measures. These thresholds helped plans to target higher quality performance on these measures. Many plans also used these targets to help incentivize provider performance. CMS stopped using the thresholds beginning in 2016. Now, CMS sets Star Ratings levels primarily based on cutpoints established annually by statistical analyses such as “clustering” of plans’ performance. In this way, plans are graded on a curve each year for most measures.

While this may encourage continual improvement efforts from plans, it makes it difficult for plans to target performance. Further, it does not guarantee that an improvement over a previous year’s performance will lead to an improved score since overall plan performance may shift. It also may make it difficult to work with providers to better drive improvements in performance.



MA contracts with a high proportion of members with social risk factors generally did worse on overall quality scores, and were much less likely to receive quality bonus payments.

Consumer Transparency

While quality measurement is used to link payments to value and thus incentivize quality improvements, it is also used to encourage beneficiaries to select higher value care. To do so, Medicare provides beneficiaries with information about the quality and value of the care they select and receive. Most commonly, CMS provides consumers information on quality via one of their “Finder” websites, such as Hospital Compare, Plan Finder, or Physician Compare. Although these sites are meant to provide consumers much needed quality and sometimes cost information, many of them are also considered to be poorly organized, not entirely consumer friendly, and lacking consistently reliable data.30 As a result, CMS continues to try to improve them.31

How quality information is presented to consumers on these websites varies. Some programs, like the MA Star Ratings, Dialysis Facility Compare and the Home Health Agency ratings, are shown using a five-Star system which combines multiple performance measures into an overall score ranging from one to five Stars. Other of CMS’ websites, particularly Physician Compare, presents a wide array of specific quality metrics but lack an easily interpreta-ble overall rating score. Further, these websites may often provide a variety of quality information to consumers, such as Hospital Finder.

Hospital Compare includes information for over 4,000 hospitals on more than 100 quality measures (though not all hospitals have information for all measures). These include an individual hospital’s scores on HRRP, HAC, and the VBP programs as well as IQR and OQR reported measures. The website also allows consumers to compare up to three hospitals at a time across the measures. CMS recently announced that it will combine 62 of the existing measures available on Hospital Compare to create a “unified” five-Star rating.32

On Physician Compare, CMS presents limited quality information, but does indicate whether physicians participate in various incentive programs such as PQRS, VPM, and EHR reporting. CMS also has made some information on quality performance for some group practices available but this is very limited.33 Additionally, ACO quality performance information, including CAHPS data, is available on the Physician Compare website.

In contrast, the MA Star Ratings have long been available to consumers and the use of the ratings for consumer information pre-dated their use for payment. CMS established the quality rating system in 2006 to help beneficiaries in their selection process for plans. Plans’ quality information is summarized in overall and measure-by-measure 5-Star ratings. CMS also uses special indicators to distinguish plans with the highest quality or that are low performing plans. One shortcoming of Plan Finder is that it presents information about FFS coverage and costs to compare with MA plans, but it does not provide comparable information on FFS quality.



Recommendations

Based on the findings presented in this comparison of Medicare quality incentive programs, several opportunities to strengthen the current MA Star Ratings program emerge. Below we briefly discuss some key recommendations for each of the four main areas we compared.

Measure Sets

CMS should reduce the number of measures included in the MA Star Ratings, while focusing on clinically meaningful outcome measures to drive greater improvements in health outcomes.

MA plans are evaluated on more measures than any other program, including ACOs. CMS should reduce the number of measures included in the MA Star Ratings to help plans better target their performance improvement efforts and resources. This change would also better align the MA Star Ratings with other Medicare value-based purchasing programs.

Also, CMS should focus on including clinically meaningful measures, especially outcome measures with a more direct link to the actions and outcomes that improve beneficiary care. In particular, we suggest that CMS attribute greater weight to data-driven outcome measures with objective clinical relevance than to measures of health status that are constructed from enrollee surveys.

CMS should continue to examine and test more meaningful adjustments to plans’ ratings to account for differences in the socioeconomic status of enrollees.

Despite the recent inclusion of an adjustment for SES through the introduc-tion of the CAI for MA plans, more needs to be done to address performance differences due to the socioeconomic differences of plans’ members. This is especially important for plans serving high proportions of beneficiaries with low SES. The CAI did not have a material impact on the overall rating for the vast majority of plans; the outcome does not warrant the complexity of the process to make the adjustment.

Given the immaterial impact of the CAI, CMS should continue actively working with the measure developers, ASPE, plans, and other stakeholders to develop and implement a long-term solution that is sufficient and meaningful. If CMS truly intends to make an impact through an adjustment aimed at addressing the impact of low-SES, it should be on par with the increments of the reward factor such that the adjustment is more meaningful and could shift more overall scores.



Scoring Methodology

Given the complexity and multitude of quality incentive programs that providers face, CMS should continue to work towards greater alignment of measures and objectives between FFS and MA plans.

In order to make a meaningful shift in quality for beneficiaries, there should be alignment across the FFS and MA programs to promote consistency and to better focus improvement efforts. Most providers see both FFS and MA members and are becoming increasingly accountable for performance over an array of metrics. In order to lessen the burden on providers as well as to reinforce areas for needed improvements, CMS should use the same or similar measures and data sources when possible. This would allow greater emphasis on overall clinical goals and would not create competing or even conflicting quality indicators for providers.

CMS should review the methodology used for the improvement measures in Part C and Part D to ensure it rewards plans for improvements in performance, protects plans that are higher performing from being penalized, and promotes improvement opportunities for all plans including lower performing plans.

Consistent with other quality incentive programs in Medicare, the calculation and the application of the improvement measures for MA plans should be improved to ensure that they do not negatively impact higher performing plans and do benefit all plans that improve. In order to protect higher performing plans, CMS should re-evaluate its “hold harmless” provision so that plans are not penalized for performance at the 4 or 5 Star level. Finally, CMS should allow plans with 2.5 Stars to benefit from the improvement measures. Limiting the measure to only plans with more than 2.5 Stars goes against the objective of the improvement measure in encouraging and rewarding improvements in performance, particularly among lower-rated plans.



CMS should reinstate four-star thresholds for selected measures to more clearly identify desired levels of performance based on prior year trends, as well as expected and realistic improvement standards, and to encourage stable improvement objectives.

CMS has previously used four star thresholds in the MA Star Ratings to help plans target their performance improvement efforts. This not only provided plans with a stable target—as opposed to one that could shift from year-to-year due to differences in performances of all plans—but also allows plans to work with providers to achieve meaningful improvements in quality. Additionally, CMS has recognized the value of performance thresholds to help drive improvement in other quality incentive programs examined.

Consumer Information

To aid Medicare beneficiaries in making a fully informed enroll-ment choice that is based on both cost and quality considerations, CMS should provide comparable measures of FFS quality for beneficiaries to consider alongside MA plans’ Star Ratings.

A key reason for evaluating and publicly releasing quality performance information is to help inform beneficiaries’ choice of plans and/or providers and to drive care toward those that are higher quality. Currently, beneficiaries are not able to make fully informed decisions with regard to quality because there is not sufficient information provided about FFS quality compared to MA plan quality. CMS should provide comparable quality information for FFS on Plan Finder so beneficiaries are better able to make informed decisions about their Medicare coverage.

While Medicare beneficiaries are able to use the Plan Finder website to compare MA plans based on their coverage, cost, and quality, Plan Finder does not include comparable quality information about FFS options.



Conclusion

The MA Star Ratings program is a successful example of effectively encouraging quality improvement though payment incentives. Despite the relative success, there continue to be aspects of the program that could be improved. By examining the MA Star Ratings in the context of other Medicare quality incentive programs, we identified specific areas that could be strengthened to better incentivize plan improvements.

Efforts to refine the measure set and methodology of calculating the MA Star Ratings could have meaningful impacts on incentivizing performance improvements. Specifically, focusing on generally reducing the number of measures while at the same time including more outcome measures that plans have a better ability to influence, providing thresholds for targeted improvements, and making more meaningful adjustments to account for plans serving higher proportions of SES are critical to continued quality improvement. It is difficult to make meaningful improvements when there are significant methodological changes each year or significant changes to the numbers and types of measures included in the programs.

Quality improvement can best happen when there are clear expectations. The recommendations presented suggest ways to enhance the MA Star Ratings so that it is more aligned with other Medicare quality incentive programs and continues to encourage plan quality improvements.



Appendix A: Overview of Pay-For-Performance Programs Examined

Program Program Description Covered Entities

Type of Incentive Program


Quality Information Available to Consumers

MA Star Ratings The MA/Part D Star Rating program assesses the quality of plans participating in the Medicare program. Plan quality is assessed over a variety of process, access and experience, and outcome (or intermediate outcome) measures. There are separate ratings for MA and Part D, but MA plans with Part D coverage receive a score for the combined set of measures. The quality ratings are only used to adjust payments to MA plans, not Part D plans.

MA and Part D Plans

Bonus payment for higher performance

5% county payment rate (10% in qualifying double bonus counties) and rebate percentage tied to Star Ratings

Yes-Plan Finder

Hospital Inpatient Quality Reporting (IQR)

The Hospital Inpatient Quality Reporting (Hospital IQR) program penalizes hospitals that do not report designated quality measures by reducing the annual update to their payment rates.

IPPS Hospitals

Pay for reporting

One-fourth reduction of annual payment update for failure to report measures

Yes-Hospital Compare

Hospital VBP The Hospital VBP program is a pay-for- performance program for inpatient hospitals that receive Medicare prospective payments. Under Hospital VBP, Medicare adjusts payments to hospitals based on either:

• How well they perform on each measure compared to all hospitals, or

• How much they improve their own perfor-mance on each measure compared to their performance during a prior baseline period.

IPPS Hospitals

Pay for performance

In 20172% withhold of base payments; funds are then redistributed to hospitals based on performance


Hospital HAC Reduces payments to IPPS hospitals that rank in the worst performing quartile of hospitals with respect to hospital-acquired conditions (HACs).

IPPS Hospitals

Penalty for poor performance

1% of total IPPS payments including add-ons


Hospital HRRP Reduces payments to acute care hospitals with excess readmissions that are paid under IPPS. The program initially focused on patients who were readmitted for selected high-cost or high-volume conditions, namely, heart attack, heart failure, and pneumonia, but now includes readmissions for COPD, elective total hip/total knee arthroplasty and coronary artery bypass grafting.

IPPS Hospitals


Up to 3% of operating payments


Hospital Outpatient Quality Reporting

The Hospital Outpatient Quality Reporting Program (Hospital OQR) is a pay for reporting program for outpatient hospital services. Requires outpatient hospitals to submit data on measures of the quality of care furnished by hospitals in outpatient settings or face reduction in OPPS payments.

OPPS Hospitals

Pay for reporting

2% reduction in OPPS payment update




Program Program Description Covered Entities

Type of Incentive Program


Quality Information Available to Consumers

PQRS The Physician Quality Reporting System (PQRS) applies a negative payment adjustment to individual eligible professionals and PQRS group practices who do not satisfactorily report data on quality measures for Medicare Part B Physician Fee Schedule (MPFS) covered professional services. Note: Physicians or practices that do not report also are penalized under the physician VBP program.

Physicians and physician practices

Pay for reporting

Total of 4%: 2%penalty for PQRS failure to report and 2% penalty under VBP

Yes-Physician Compare

Physician VBPM The Value Modifier provides for differential payment to a physician or group of physicians under the Medicare Physician Fee Schedule based on the quality of care furnished com-pared to the cost of care during a performance period. The Value Modifier is used to adjust Medicare payments to non-physician eligible professionals, in addition to physicians.


Pay for performance

In 2017: For groups of 10-99 practitioners adjustment of -2% to +2%. For groups of 100 or more: adjustment of -4% to +4%


Meaningful Use Meaningful Use is a payment adjustment for eligible providers based on their use of electronic health records (EHR).


Penalty (incentive payments were used until 2015)

3≥% in 2017(shift to MIPS in 2018)


MIPS Merit-Based Incentive Payment System (MIPS) will replace PQRS, physician VBPM, and meaningful use beginning in 2019. Physicians will receive positive or negative adjustments depending on performance across a range of categories including: Quality, Resource Use, Advancing Care Information, and Clinical Practice Improvement Activities.


Pay for performance

Will impact 4% of payments in 2019 and up to 9% by 2022 (high-performing providers can receive higher payment updates

TBD

ESRD QIP The ESRD QIP reduces payments to ESRD facilities that do not meet certain performance standards..

IPPS Hospitals


Up to 3% of operating payments

Yes-Dialysis Facilities Compare

HHA Quality Reporting

Home health agencies (HHAs) are required to submit quality data, and payments are reduced for agencies that do not submit the information.

Home Health Agencies

Pay for reporting

2% penalty for failure to report

Yes-Home Health Compare

HHA VBP Pilot Program

Pilot program for home health agencies in 9 states that ties Medicare payments to quality performance. The pilot runs from January 1,2016through December 31, 2022. 2016 is the first performance year; payment implications begin in 2018.

Home Health Agencies (in 9 states)

Pay for performance

Bonus or penalty payments of 5-8% beginning in 2018 depending on performance

NA

ACO Quality Measure Program

In order to receive any shared savings amounts, ACOs must also meet certain quality performance standards.

ACOs Shared Savings Eligibility

Determines shared savings availability


SNF VBP (begins in FY 2019)

SNFs will have payment tied to performance based on either achievement or improvement.

Skilled Nursing Facilities

Penalty Up to 2% of Part A payments

Yes-To be posted on Nursing Home Compare



Endnotes

1 James, J. (2012). “Health Policy Brief: Pay-for-Performance,” Health Affairs. Retrieved from http://www.rwjf.org/content/dam/farm/reports/issue_briefs/2012/rwjf401782 (accessed April 10, 2017).

2 Department of Health and Human Services, Centers for Medicare & Medicaid Services, Medicare Learning Net-work. (2015). Hospital Value-Based Purchasing. Retrieved from https://www.cms.gov/Outreach-and-Education/Medicare-Learning-Network-MLN/MLNProducts/downloads/Hospital_VBPurchasing_Fact_Sheet_ICN907664.pdf (accessed April 11, 2017).

3 Centers for Medicare & Medicaid Services. (Accessed 2017). Value-Based Payment Modifier. Retrieved from https://www.cms.gov/medicare/medicare-fee-for-service-payment/physicianfeedbackprogram/valuebasedpay-mentmodifier.html (accessed April 11, 2017).

4 American Academy of Family Physicians (AAFP). (Accessed 2017). MACRA Basics FAQ on MACRA and Medicare Payment Reform. Retrieved from http://www.aafp.org/practice-management/payment/medicare-payment/ faq.html (accessed April 11, 2017).

5 Kahn,C.N., Ault, T., Potetz, L., Walke, T., Chambers, J.H., & Burch, S. (2015). Assessing Medicare’s Hospital Pay-For-Performance Programs And Whether They Are Achieving Their Goals. Health Affairs, 2015;34(8):1281–8. Retrieved from http://content.healthaffairs.org/content/34/8/1281.abstract (accessed April 11, 2017).

6 Centers for Medicare & Medicaid Services. (2016). Medicare 2017 Part C & D Star Rating Technical Notes. Re-trieved from https://www.cms.gov/medicare/prescription-drug-coverage/prescriptiondrugcovgenin/ performancedata.html (accessed April 12, 2017).

7 Ibid.

8 AIS Health Presentation. (November, 2016). The 2017 Medicare Star Ratings: How to Translate New CMS Data Into Future Successes (with analyses by Inovalon). Retrieved from http://resources.inovalon.com/webinars/ inovalon-webinar-2017-cms-star-ratings-and-quality-improvement-strategies (accessed July 12, 2017).

9 See for example Cassel, C., Conway, P., Delbanco, S., Jha, A., Saunders, R., & Lee,T. (2014). Getting More Perfor-mance from Performance Measurement, N Engl J Med; 371:2145-2147, DOI: 10.1056/NEJMp1408345; and Kenen, J. (2015). Look at How Health Quality Measures Have Become a Jungle. Retrieved from http://healthjournalism.org/blog/2015/08/look-at-how-health-quality-measures-have-become-a-jungle/ (accessed April 11, 2017).

10 The Lake Superior Quality Innovation Network serves Michigan, Minnesota, and Wisconsin, under the Centers for Medicare & Medicaid Services Quality Improvement Organization Program. (updated 2015). Understanding Value-Based Purchasing. Retrieved from https://www.lsqin.org/wp-content/uploads/2015/11/FY2018-VBP- Fact-Sheet-11.10.pdf (accessed April 11, 2017).

11 Government Accountability Office (GAO). (2015). Hospital Value-Based Purchasing: Initial Results Show Modest Effects on Medicare Payments and No Apparent Change in Quality-of-Care Trends. Retrieved from http://www.gao.gov/assets/680/672899.pdf (accessed April 11, 2017).

12 Centers for Medicare & Medicaid Services. (2008).Roadmap for Quality Measurement in the Traditional Medicare Fee-for-Service Program. Retrieved from https://www.cms.gov/Medicare/Quality-Initiatives- Patient-Assessment-Instruments/QualityInitiativesGenInfo/downloads/QualityMeasurementRoadmap_ OEA116_508.pdf (accessed April 11, 2017).

13 National Quality Forum. (2014). Risk Adjustment for Socioeconomic Status or Other Sociodemographic Factors. Retrieved from http://www.qualityforum.org/Publications/2014/08/Risk_Adjustment_for_Socioeconomic_ Status_or_Other_Sociodemographic_Factors.aspx (accessed April 11, 2017).

14 Office of the Assistant Secretary for Planning and Evaluation (ASPE). (2016). Report to Congress: Social Risk Factors and Performance Under Medicare’s Value-Based Payment Programs. Retrieved from http://www.quality-forum.org/Publications/2014/08/Risk_Adjustment_for_Socioeconomic_Status_or_Other_Sociodemographic_ Factors.aspx (accessed April 11, 2017).

15 Ibid.

16 Reidhead, M. Hospital Industry Data Institute, HealthStats, Missouri Hospital Association. (2016). Including Sociodemographic Factors in Risk-Adjusted Readmission Measures. Retrieved from http://www.mhanet.com/mhaimages/SociodemFactos_HealthStats_0216.pdf (accessed April 11, 2017).

17 See Hu, J., Gonsahn, M. D., & Nerenz, D. R. (2014). Socioeconomic status and readmissions: evidence from an urban teaching hospital. Health Affairs, 33(5), 778-785; and Calvillo–King, L., Arnold, D., Eubank, K. J., Lo, M., Yunyongying, P., Stieglitz, H., & Halm, E. A. (2013). Impact of social factors on risk of readmission or mortality in pneumonia and heart failure: systematic review. Journal of general internal medicine, 28(2), 269-282.

18 See Berenson, J. & Shih, A. (2012). Higher Readmissions at Safety-Net Hospitals and Potential Policy Solutions. Commonwealth Fund. Retrieved from https://www.commonwealthfund.org/publications/issue-briefs/2012/dec/higher-readmissions-safety-net-hospitals-and-potential-policy (accessed April 27, 2017); and Barnett, M.L., Hsu, J., & McWilliams, J.M.(2015). Patient Characteristics and Differences in Hospital Readmission Rates. JAMA Internal Medicine.

19 Gilman, E., Adams, E.K., Hockberry, J.M., Milstein, A.S., Wilson, I.B., & Becker, E.R. (2015). Safety-Net Hospitals More Likely than Other Hospitals to Fare Poorly under Medicare’s Value-Based Purchasing. Health Aff (Millwood). 2015; 34(3):398–405.

http://www.rwjf.org/content/dam/farm/reports/issue_briefs/2012/rwjf401782

http://www.rwjf.org/content/dam/farm/reports/issue_briefs/2012/rwjf401782

https://www.cms.gov/Outreach-and-Education/Medicare-Learning-Network-MLN/MLNProducts/downloads/Hospital_VBPurchasing_Fact_Sheet_ICN907664.pdf



https://www.cms.gov/medicare/medicare-fee-for-service-payment/physicianfeedbackprogram/valuebasedpaymentmodifier.html

https://www.cms.gov/medicare/medicare-fee-for-service-payment/physicianfeedbackprogram/valuebasedpaymentmodifier.html

http://www.aafp.org/practice-management/payment/medicare-payment/faq.html

http://www.aafp.org/practice-management/payment/medicare-payment/faq.html

http://content.healthaffairs.org/content/34/8/1281.abstract

https://www.cms.gov/medicare/prescription-drug-coverage/prescriptiondrugcovgenin/performancedata.html


http://resources.inovalon.com/webinars/inovalon-webinar-2017-cms-star-ratings-and-quality-improvement-strategies

http://resources.inovalon.com/webinars/inovalon-webinar-2017-cms-star-ratings-and-quality-improvement-strategies

http://healthjournalism.org/blog/2015/08/look-at-how-health-quality-measures-have-become-a-jungle/

http://healthjournalism.org/blog/2015/08/look-at-how-health-quality-measures-have-become-a-jungle/

https://www.lsqin.org/wp-content/uploads/2015/11/FY2018-VBP-Fact-Sheet-11.10.pdf

https://www.lsqin.org/wp-content/uploads/2015/11/FY2018-VBP-Fact-Sheet-11.10.pdf

http://www.gao.gov/assets/680/672899.pdf

https://www.cms.gov/Medicare/Quality-Initiatives-Patient-Assessment-Instruments/QualityInitiativesGenInfo/downloads/QualityMeasurementRoadmap_OEA1-16_508.pdf



http://www.qualityforum.org/Publications/2014/08/Risk_Adjustment_for_Socioeconomic_Status_or_Other_Sociodemographic_Factors.asp

http://www.qualityforum.org/Publications/2014/08/Risk_Adjustment_for_Socioeconomic_Status_or_Other_Sociodemographic_Factors.asp

http://www.qualityforum.org/Publications/2014/08/Risk_Adjustment_for_Socioeconomic_Status_or_Other_Sociodemographic_Factors.aspx



http://www.mhanet.com/mhaimages/SociodemFactos_HealthStats_0216.pdf

http://www.mhanet.com/mhaimages/SociodemFactos_HealthStats_0216.pdf

https://www.commonwealthfund.org/publications/issue-briefs/2012/dec/higher-readmissions-safety-net-hospitals-and-potential-policy

https://www.commonwealthfund.org/publications/issue-briefs/2012/dec/higher-readmissions-safety-net-hospitals-and-potential-policy



20 Kahn, C.N., Ault, T., Potetz, L., Walke, T., Chambers, J.H., and Burch, S. (2015). Assessing Medicare’s Hospital Pay-For-Performance Programs And Whether They Are Achieving Their Goals. Health Affairs, 2015; 34(8): 1281–8. Retrieved from https://www.healthaffairs.org/doi/abs/10.1377/hlthaff.2015.0158 (accessed April 11, 2017).

21 National Quality Forum. (2016). Final Report: MAP 2016 Considerations for Implementing Measures in Federal Programs: Clinicians. Retrieved from http://www.qualityforum.org/Publications/2016/03/MAP_2016_Consider-ations_for_Implementing_Measures_in_Federal_Programs__Clinicians.aspx (accessed April 12, 2017).

22 National Quality Forum. (2016). NQF Provides Measure Recommendations to HHS for MIPS and Other Federal Healthcare Programs. Retrieved from http://www.qualityforum.org/News_And_Resources/Press_Releas-es/2016/NQF_Provides_Measure_Recommendations_to_HHS_for_MIPS_and_Other_Federal_Healthcare_Pro-grams.aspx (accessed April 12, 2017).

23 See for example Weiss, H., & Pescatello, S. (2014). Medicare Advantage: Stars System’s Disproportionate Impact On MA Plans Focusing On Low-Income Populations. Retrieved from https://www.healthaffairs.org/do/10.1377/hblog20140922.041500/full/ (accessed April 12, 2017); and Inovalon Blog. (2015). Inovalon Study Demonstrates Need for Star Rating Adjustments. Retrieved from http://www.inovalon.com/inovalon-in-sights-blog/inovalon-study-demonstrates-need-star-rating-adjustments

24 Centers for Medicare & Medicaid Services. (2015). Examining the Potential Effects of Socioeconomic Factors on Star Ratings. Slide presentation on RAND Study Results. Retrieved from https://www.cms.gov/medicare/prescription-drug-coverage/prescriptiondrugcovgenin/performancedata.html (accessed April 12, 2017).

25 Centers for Medicare & Medicaid Services. (2016). Announcement of Calendar Year (CY) 2017 Medicare Advantage Capitation Rates and Medicare Advantage and Part D Payment Policies and Final Call Letter. Retrieved from https://www.cms.gov/Medicare/Health-Plans/MedicareAdvtgSpecRateStats/Announcements- and-Documents-Items/2017Anouncement.html?DLPage=1&DLEntries=10&DLSort=2&DLSortDir=descending (accessed December 17, 2016).

26 Centers for Medicare & Medicaid Services. (2015). 2013 Reporting Experience Including Trends (2007-2014): Physician Quality Reporting System and Electronic Prescribing (eRx) Incentive Program. Retrieved from https://www.facs.org/~/media/files/advocacy/regulatory/2013_pqrserx_experience_report.ashx (accessed April 11, 2017).

27 See for example Rice, S. (2015). Physician Quality Pay Not Paying Off. Modern Healthcare. Retrieved from http://www.modernhealthcare.com/article/20150530/magazine/305309979 (accessed April 11, 2017).

28 See for example American College of Physicians Comment Letter on MACRA Proposed Rule (2016) retrieved from https://www.acponline.org/acp_policy/letters/comment_letter_macra_proposed_rule_2016.pdf (accessed April 11, 2017).

29 Advisory Board, Care Transformation Center Blog. (2016). The ACO Quality Challenge: Not Just About Reporting. Retrieved from https://www.advisory.com/research/care-transformation-center/care-transformation- center-blog/2016/01/cqr-not-just-reporting (accessed April 11, 2017).

30 Government Accountability Office (GAO), HEALTH CARE TRANSPARENCY: Actions Needed to Improve Cost and Quality Information for Consumers, GAO-15-11: Published: Oct 20, 2014. Publicly Released: Nov 18, 2014. Retrieved from http://www.gao.gov/products/GAO-15-11 (accessed July 10, 2017).

31 Findlay, S. (updated 2015). Physician Compare. Health Policy Briefs, Health Affairs. Retrieved from http://www.healthaffairs.org/healthpolicybriefs/brief.php?brief_id=146 (accessed April 11, 2017).

32 Centers for Medicare & Medicaid Services, Fact Sheet. (2016). Data Brief: Evaluation of National Distributions of Overall Hospital Quality Star Ratings. https://www.cms.gov/Newsroom/MediaReleaseDatabase/Fact-sheets/2016-Fact-sheets-items/2016-07-21-2.html (accessed April 11, 2017). Centers for Medicare & Medicaid Services. (Accessed 2017). Quality Data and Physician Compare. Retrieved from https://www.cms.gov/ Medicare/Quality-Initiatives-Patient-Assessment-Instruments/physician-compare-initiative/ Quality-Data-and-Physician-Compare-.html#What%20quality%20measures%20are%20currently%20on%20Physician%20Compare? (accessed April 11, 2017).

https://www.healthaffairs.org/doi/abs/10.1377/hlthaff.2015.0158

http://www.qualityforum.org/Publications/2016/03/MAP_2016_Considerations_for_Implementing_Measures_in_Federal_Programs__Clinicians.aspx

http://www.qualityforum.org/Publications/2016/03/MAP_2016_Considerations_for_Implementing_Measures_in_Federal_Programs__Clinicians.aspx

http://www.qualityforum.org/News_And_Resources/Press_Releases/2016/NQF_Provides_Measure_Recommendations_to_HHS_for_MIPS_and_Other_Federal_Healthcare_Programs.aspx



https://www.healthaffairs.org/do/10.1377/hblog20140922.041500/full/

https://www.healthaffairs.org/do/10.1377/hblog20140922.041500/full/

http://www.inovalon.com/inovalon-insights-blog/inovalon-study-demonstrates-need-star-rating-adjustments

http://www.inovalon.com/inovalon-insights-blog/inovalon-study-demonstrates-need-star-rating-adjustments



https://www.cms.gov/Medicare/Health-Plans/MedicareAdvtgSpecRateStats/Announcements-and-Documents-Items/2017Anouncement.html?DLPage=1&DLEntries=10&DLSort=2&DLSortDir=descending

https://www.cms.gov/Medicare/Health-Plans/MedicareAdvtgSpecRateStats/Announcements-and-Documents-Items/2017Anouncement.html?DLPage=1&DLEntries=10&DLSort=2&DLSortDir=descending

https://www.facs.org/~/media/files/advocacy/regulatory/2013_pqrserx_experience_report.ashx

http://www.modernhealthcare.com/article/20150530/magazine/305309979

https://www.acponline.org/acp_policy/letters/comment_letter_macra_proposed_rule_2016.pdf

https://www.advisory.com/research/care-transformation-center/care-transformation-center-blog/2016/01/cqr-not-just-reporting

https://www.advisory.com/research/care-transformation-center/care-transformation-center-blog/2016/01/cqr-not-just-reporting

http://www.gao.gov/products/GAO-15-11

https://www.cms.gov/Newsroom/MediaReleaseDatabase/Fact-sheets/2016-Fact-sheets-items/2016-07-21-2.ht

https://www.cms.gov/Newsroom/MediaReleaseDatabase/Fact-sheets/2016-Fact-sheets-items/2016-07-21-2.ht

https://www.cms.gov/Medicare/Quality-Initiatives-Patient-Assessment-Instruments/physician-compare-initiative/Quality-Data-and-Physician-Compare-.html#What%20quality%20measures%20are%20currently%20on%20Physician%20Compare?




Anthem Public Policy Institute

Anthem’s Public Policy Institute (PPI) was established to share data and insights to inform public policy and shape the healthcare programs of the future. The Public Policy Institute strives to be an objective and credible contributor to healthcare innovation and transformation through publication of policy-relevant data analysis, timely research, and insights from Anthem’s innovative programs.

AnthemPublicPolicyInstitute.com

ABOUT US

Anthem, Inc.

Anthem is working to transform healthcare with trusted and caring solutions. Our health plan companies deliver quality products and services that give consumers access to the care they need. With over 73 million people served by its affiliated companies, including more than 40 million enrolled in its family of health plans, Anthem is one of the nation’s leading health benefits companies.

AnthemInc.com


http://AnthemPublicPolicyInstitute.com

http://AnthemInc.com